Open X-Embodiment

license

CC-BY-4.0(all non-software materials, per the README)

access

public(TFDS and a public GCS bucket)

dataset_card

present(README, dataset spreadsheet and paper)

The whole collection downloads without an account under CC BY 4.0, and the paper and README document its format and sources.

• https://arxiv.org/abs/2310.08864 recorded 2026-09-26

  arXiv 2310.08864, "Open X-Embodiment: Robotic Learning Datasets and RT-X Models", documents the collection and its standardized format.

• https://raw.githubusercontent.com/google-deepmind/open_x_embodiment/HEAD/README.md recorded 2026-09-26

  README: "All software is licensed under the Apache License, Version 2.0 ... All other materials are licensed under the Creative Commons Attribution 4.0 International License (CC-BY)"; datasets load with tfds.load or gsutil from gs://gdm-robotics-open-x-embodiment.

GitHub stars on the dataset's repository. The data is served from TensorFlow Datasets and a Cloud Storage bucket, neither of which publishes download counts, and the Hub copies of member datasets are other people's conversions.

• https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/google-deepmind%2Fopen_x_embodiment recorded 2026-09-26

  Repository record for google-deepmind/open_x_embodiment: 2,045 stargazers.

Open X-Embodiment pools more robot bodies than any other collection here, from labs around the world, and it is the dataset most of the open robot policies in this category were trained on.

• https://robotics-transformer-x.github.io/ recorded 2026-09-26

  Project page: "1M+ real robot trajectories spanning 22 robot embodiments, from single robot arms to bi-manual robots"; "a dataset from 22 different robots collected through a collaboration between 21 institutions, demonstrating 527 skills (160266 tasks)".